The Bulge Asymmetries and Dynamical Evolution (Baade) Sio Maser Survey at 86 Ghz with ALMA

DRAFT VERSION SEPTEMBER 6, 2019 Typeset using LATEX twocolumn style in AASTeX62 The Bulge Asymmetries and Dynamical Evolution (BAaDE) SiO Maser Survey at 86 GHz with ALMA ∗ MICHAEL C. STROH,1 YLVA M. PIHLSTRÖM,1 , LORÁNT O. SJOUWERMAN,2 MEGAN O. LEWIS,1 MARK JCLAUSSEN,2 MARK R. MORRIS,3 AND R. MICHAEL RICH3 1Department of Physics & Astronomy, The University of New Mexico, Albuquerque, NM 87131 2National Radio Astronomy Observatory, Array Operations Center, Socorro, NM 87801 3Department of Physics & Astronomy, University of California, Los Angeles, CA 90095 (Accepted August 14th, 2019) ABSTRACT We report on the first 1,432 sources observed using the Atacama Large Millimeter/submillimeter Array (ALMA), from the Bulge Asymmetries and Dynamical Evolution (BAaDE) survey, which aims to obtain tens of thousands of line-of-sight velocities from SiO masers in Asymptotic Giant Branch (AGB) stars in the Milky Way. A 71% detection rate of 86 GHz SiO masers is obtained from the infrared color-selected sample, and increases to 80% when considering the likely oxygen-rich stars using Midcourse Space Experiment (MSX) colors isolated in a region where [D] - [E] ≤ 1:38. Based on Galactic distributions, the presence of extended CS emission, and likely kinematic associations, the population of sources with [D] - [E] > 1:38 probably consists of young stellar objects, or alternatively, planetary nebulae. For the SiO detections we examined whether indi- vidual SiO transitions provide comparable stellar line-of-sight velocities, and found that any SiO transition is suitable for determining a stellar AGB line-of-sight velocity. Finally, we discuss the relative SiO detection rates and line strengths in the context of current pumping models. Keywords: masers — stars: infrared — stars: late-type — radio lines: stars — surveys — Galaxy: center — Galaxy: kinematics and dynamics 1. INTRODUCTION the measured gravitational potential. Thus a large and uni- The Bulge Asymmetries and Dynamical Evolution (BAaDE) form stellar survey is required to complement the large scale survey aims to improve our understanding of the structure of gas surveys. the inner Galaxy and Galactic Bulge (L. O. Sjouwerman et Infrared Astronomical Satellite (IRAS) two-color dia- al. 2019, in preparation). By using line-of-sight velocities of grams have successfully differentiated between oxygen- and SiO maser emission from red giant stars, these stars act as carbon-rich, evolved Asymptotic Giant Branch (AGB) pop- point-like probes of the Galactic gravitational potential. ulations, while also distinguishing between less evolved ob- Interstellar extinction greatly reduces the extent of the jects with thinner envelopes (Miras and semi-regular vari- Milky Way that surveys utilizing stellar or compact probes ables) and the thicker envelopes associated with OH/IR stars (van der Veen & Habing 1988). However, the 10 angular reso- arXiv:1909.02090v1 [astro-ph.GA] 4 Sep 2019 can reach, in turn necessitating piecemeal approaches to model the spiral arms and Galactic bulge. Since the spiral lution of IRAS leads to confusion close to the Galactic plane, arms span large ranges of Galactic longitude, to constrain the limiting its usefulness for source selection in regions clos- Galactic structure models, multiple surveys must be used, of- est to the Milky Way’s disk. Midcourse Space Experiment 00 ten with their own limitations. Gas emission has been much (MSX) has an 18 point spread function, thus it does not suf- more successful in uniformly mapping larger regions of the fer from the same level of source confusion throughout the Milky Way(e.g. the CO(1 - 0) survey by Dame et al. 2001); Galactic plane as IRAS. Differentiating between these nu- however, gas and stellar probes may lead to differences in merous populations is more difficult with MSX mid-infrared colors, as the longest wavelength band is only 21.3µm, compared to the longer 60µm band of IRAS. However, substan- ∗ Y. M. Pihlström is also an Adjunct Astronomer at the National Radio Astronomy Observatory. 2 M.C. STROH ET AL. tial progress has been made supplementing MSX colors with Table 1. ALMA spectral window configuration data in the near-infrared (see Section 4.1.1). Central Channel Number Potential The BAaDE survey assumes sampling of an oxygen-rich, Mira-dominated population since it is associated with region Spectral Frequency Width of Lines iiia of the MSX two-color diagram, thus consisting of stars Window (MHz) (km/s) Channels Covered with thin circumstellar shells providing conditions favorable 2A 85 646.6055 0.855 1920 SiO v=2, 29SiO v=0 for SiO maser emission (Sjouwerman et al. 2009). Addi- 1 86 249.6630 0.848 3840 SiO v=1, H13CN tionally, both radiative and collisional excitation models sug- 2B 86 853.2095 0.843 1920 SiO v=0, H13CO+ gest that SiO maser emission forms in the shells around thin- 3 97 988.0684 0.747 3840 CS shelled AGB stars, with the maser emission radiating tangen- NOTE—The channel width is calculated at the central frequency in each spec- tial to the shell surface. The expected ring-like structure of tral window. the maser emission regions has been confirmed by Very Long Baseline Interferometry (VLBI) observations (e.g. Miyoshi et al. 1994; Desmurs et al. 2000; Diamond & Kemball 2003), and demonstrates that SiO maser emission is usually an ac- curate tracer of the stellar line-of-sight velocity. Jewell et al.(1991) found no systematic red- or blue-shifting of the SiO v=1 emission velocity compared to the velocities estab- lished by thermal CO and OH emission, further validating the strength of SiO maser emission as a stellar line-of-sight velocity tracer. The full BAaDE sample consists of 28,062 sources selected by MSX colors. All sources were selected from version 2.3 of the MSX point source catalog (Egan et al. 2003) that coincide with MSX color region iiia (Sjouwerman et al. 2009). The MSX A, C, D and E bands are centered on wave- lengths of 8.28, 12.13, 14.65 and 21.34µm, respectively1. MSX iiia color boundaries are given by -0:6 ≤ [A] - [D] ≤ +0:4, or -0:7 ≤ [A] - [E] ≤ +0:4 or -0:75 ≤ [C] - [E] ≤ Figure 1. Spectral windows for BAaDE ALMA observations. The +0:1. The MSX [A]–[D], [A]–[E], and [C]–[E] color selec- black lines represent the rest frequency of each transition, with the -1 tions correspond to 27,130 candidates, 563 candidates, and gray regions representing a ±400 km s range surrounding each 369 candidates, respectively. The majority (18,988) of the rest frequency. BAaDE sources have been surveyed with the Karl G. Jan- sky Very Large Array (VLA) using the 43 GHz (J=1–0) ro- is instead surveyed in the J=2–1 rotational transitions at 86 tational transitions of SiO. The VLA cannot observe 9,074 GHz. BAaDE sources (32% of the full sample) since they are lo- Section2 describes the observations and spectral line de- cated at declinations below -35◦. This low-declination sam- tection algorithm. Section3 summarizes the line detection ple at -110◦ < l < -5◦ contains regions of the Galactic bar results. Section4 discusses the CS emission population, how furthest from us. Some symmetry along the Galactic bar and the majority of CS emitters are likely not associated with late- the spiral structure may be assumed when modeling the struc- type AGB stars, and how they can be filtered out of the sur- ture of the Galactic plane, however, filling in this region with vey. Finally, Section5 discusses general trends of the 86 GHz data from stellar objects using ALMA is essential for metic- SiO masers and how they relate to pumping models. ulous testing of the structure of the Milky Way when com- 2. OBSERVATIONS AND DATA ANALYSIS bined with the larger BAaDE sample observed with the VLA. Since ALMA cannot observe the 43 GHz frequencies asso- This paper reports on the 1432 sources observed during ciated with J=1–0 SiO maser transitions, the ALMA sample ALMA Cycles 2, 3 and 5 (June 2014 through September 2018). To minimize slewing overhead and enable efficient phase calibration, the sources were grouped based on their 1 Henceforth, A, C, D & E will refer to the MSX bands, while [A], [C], angular separation. [D] & [E] will refer to the zero-magnitude corrected magnitudes for the respective band. For example [D] - [E] refers to the difference between the 2.1. Observations zero-magnitude corrected magnitudes of the D & E MSX bands. The ALMA spectral setup in Band 3 is described in Table 1 and in Figure1. The main objective was to cover the SiO THE BAADE SURVEY AT 86 GHZ USING ALMA 3 v=0, 1 & 2 transitions, which also allowed coverage of the will require significantly more computer and time resources 29SiO v=0 isotopologue transition. Unless otherwise noted, than currently available. Pihlström et al.(2018) found that the SiO transitions refer to the 86 GHz J=2–1 transitions. for known 43 GHz SiO masers, only a 0:0012 mean offset ex- Additionally, SiO refers to the 28SiO transitions and we will ists between the derived VLA and 2MASS positions. Since specifically reference 29SiO when discussing isotopologue 2MASS associations exist for 96% of the BAaDE survey, in transitions. With the aim of deriving velocities from sources order to increase our detection rate, if a 2MASS association that instead may belong to a carbon-rich population, an ad- existed for a BAaDE source, the phase center was shifted to ditional baseband was placed on the CS (J=2–1) transition the 2MASS position using the task fixvis. The phase center at 98 GHz, which has been observed, for example, in the was adjusted before any standard calibration was performed.

Load more